What role do data network effects play for multihoming complements? An empirical analysis of platform complementors’ strategies for consumer value creation

Classified as probably not a substantial claim.

Leveraging data for value creation in digital platform ecosystems is relevant for the strategic choices of complementors, such as their multihoming strategy, platform product design decisions regarding software and hardware autonomy, or the type of value capture.

“leveraging data for value creation in digital platform ecosystems is relevant for the strategic choices of complementors (Cennamo, 2018;Chen et al., 2022), such as their multihoming strategy, platform product design decisions (Gregory et al., 2021) regarding software and hardware autonomy (Hein et al., 2020), or the type of value capture of platform participants”

• cites Cennamo, 2018 sourceNot checked — no full text of the source
• cites Chen et al., 2022 sourcePartly supported

  Split into 2 atoms — 1 confirmed, 1 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

  Experimental — accuracy not yet measured
• cites Gregory et al., 2021 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Data variety, updates, and extensions enable AI models to improve continuously and avoid biases or drifts.

“In turn, however, data variety, updates, and extensions enable AI models to improve continuously and avoid biases or drifts, such as imprecise results due to obsolete data (Janiesch et al., 2021).”

• cites Janiesch et al., 2021 sourcePartly supported

  In “Awareness of bias and drift in data” (p. 7), the paper says obsolete historical data can impair results and that periodic retraining is used to counter concept drift. It does not say that data variety, updates, and extensions continuously improve models or prevent bias and drift; it describes bias as inherent in data-based applications and drift as constant in real-world electronic markets. (LLM, conf=0.87)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Due to the costs of innovation in complex digital platforms, complements perceived as high quality in one platform are not necessarily candidates for a multihoming strategy.

“However, due to the costs of innovation in complex digital platforms (Cennamo et al., 2018), complements perceived as high quality in one platform are not necessarily candidates for a multihoming strategy.”

• cites Cennamo et al., 2018 sourceSupported

Probably not a substantial claim: may not pass S3 (load-bearing).

iOS developers act in a mature ecosystem, characterized by competitive pressure on the complementor side.

“In addition, iOS developers act in a mature ecosystem, characterized by competitive pressure on the complementor side (Cenamor, 2021;de Reuver et al., 2018).”

• cites Cenamor, 2021 sourceNot checked — no full text of the source
• cites de Reuver et al., 2018 sourcePartly supported

  Section 3.3 (page unavailable) describes app developers publishing on iOS and Android and participating in overlapping platform ecosystems. It says platforms compete with other platforms, but does not characterize the iOS ecosystem as mature or describe competitive pressure among complementors. (LLM, conf=0.76)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

Consumer value is multifaceted and susceptible to various factors, such as consumer characteristics, the offering under consideration, and the specific situation.

“Consumer value is multifaceted and susceptible to various factors, such as consumer characteristics, the offering under consideration, and the specific situation (Henfridsson et al., 2018;Suseno et al., 2018).”

• cites Henfridsson et al., 2018 sourcePartly supported

  “The Two Sides of Recombination” says use value differs among users and depends on which resources in an offering they value. “Outcomes” defines value in relation to a user in a particular context. These passages support variation by user, offering, and context for digital innovations, but do not establish the broader claim about consumer value generally or identify consumer characteristics as a factor. (LLM, conf=0.82)

  Experimental — accuracy not yet measured
• cites Suseno et al., 2018 sourceSupported

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing). May combine two claims.

Platform users' consumption defines the value, suggesting that value is created in use rather than in exchange.

“Ecosystem research emphasizes that platform users' consumption defines the value (Autio & Thomas, 2018;Henfridsson et al., 2018), suggesting that value is created”

• cites Autio & Thomas, 2018 sourceNot checked — no full text of the source
• cites Henfridsson et al., 2018 sourceSupported
• cites Lusch & Nambisan, 2015 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S2 (checkable).

Due to network interaction, individual actors' choices affect other actors' value.

“Due to network interaction, individual actors' choices affect other actors' value (Jacobides et al., 2024;Rochet & Tirole, 2003)”

• cites Jacobides et al., 2024 sourceSupported
• cites Rochet & Tirole, 2003 sourceSupported

Probably not a substantial claim: may not pass S3 (load-bearing).

Complementors can decide to approach new market opportunities, particularly when encountering strong competition within a market.

“Beyond those strategic considerations within one platform ecosystem, complementors can also decide to approach new market opportunities, particularly when encountering strong competition within a market (Corts & Lederman, 2009;Nerbel & Kreutzer, 2023).”

• cites Corts & Lederman, 2009 sourcePartly supported

  Section II (Industry Background; page unavailable) says game publishers may release titles on multiple platforms to reach a larger audience. Section III.B (Software supply; page unavailable) discusses how growth in rival platforms can draw software development away from a focal platform. These passages support pursuing users on other platforms, but do not establish that complementors seek new markets particularly because they face strong competition within a market. (LLM, conf=0.82)

  Experimental — accuracy not yet measured
• cites Nerbel & Kreutzer, 2023 sourceSupported

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

Multihoming complementors must determine a strategic balance in their resource investments to still serve consumer interests.

“Hence, they must determine a strategic balance in their resource investments to still serve consumer interests (Kapoor & Agarwal, 2017).”

• cites Kapoor & Agarwal, 2017 sourcePartly supported

  The Additional information section (page unavailable) says that participation in both iOS and Android may create resource-allocation challenges, supporting the multihoming and investment-balance aspect. It does not say that complementors must strike that balance to serve consumer interests; the paper frames the issue around sustaining complementors’ performance. (LLM, conf=0.84)

  Experimental — accuracy not yet measured

Classified as probably not a substantial claim.

Data can be continuously processed and adapted to new contexts without diminishing as they are editable, portable, and recontextualizable.

“Data can be continuously processed and adapted to new contexts without diminishing as they are editable, portable, and recontextualizable (Aaltonen et al., 2021).”

• cites Aaltonen et al., 2021 sourcePartly supported

  The Literature Review describes data as editable, portable, and recontextualizable, and Becoming a Data Commodity describes their iterative transformation and repurposing. But the paper does not establish that this happens without diminishing them; that section says data can gain and lose properties, while The Production of Data Tokens notes that editability can lead to loss of relevance. (LLM, conf=0.86)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

Value creation for consumers occurs when the value produced by individual actors' efforts exceeds the value complementors capture.

“Value creation for consumers occurs when the value produced by individual actors' efforts exceeds the value complementors capture (Jacobides et al., 2024;Lepak et al., 2007).”

• cites Jacobides et al., 2024 sourcePartly supported

  Section 3.3 describes actors’ joint value production and value propositions for end customers. It does not say that consumer value creation occurs when the value produced by individual actors exceeds what complementors capture; section 4.3 instead discusses how value-capture tensions can discourage investment and reduce value delivered to customers. Page numbers are unavailable in our copy. (LLM, conf=0.82)

  Experimental — accuracy not yet measured
• cites Lepak et al., 2007 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

The monetization of complementors' products might affect consumer satisfaction.

“Complementors aim for high-quality outputs to meet users' expectations; however, the monetization of their products might affect consumer satisfaction (Cennamo et al., 2018;Clough & Wu, 2022).”

• cites Cennamo et al., 2018 sourceCould not verify

  Section 3.1 says users assess complements’ quality, while the Introduction discusses the revenue incentive to develop for multiple platforms. Neither passage links product monetization to consumer satisfaction. Searches for monetization, pricing, and satisfaction found no passage making that connection. (LLM, conf=0.75)

  Experimental — accuracy not yet measured
• cites Clough & Wu, 2022 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

In the context of multihoming, research suggests the specialization of complements toward a particular platform ecosystem potentially has a positive effect on complement quality.

“In the context of multihoming, research suggests the specialization of complements toward a particular platform ecosystem potentially has a positive effect on complement quality (Cennamo et al., 2018;Kapoor & Agarwal, 2017).”

• cites Cennamo et al., 2018 sourceSupported
• cites Kapoor & Agarwal, 2017 sourcePartly supported

  The Ecosystem Experience section links accumulated development experience to building a high-quality app, and Additional information notes that participation in both ecosystems may create resource-allocation challenges. However, the paper does not establish that specializing in one platform, rather than multihoming, improves complement quality; its main outcome is sustained superior performance, measured by app revenue ranking. The indexed passages provide no page numbers. (LLM, conf=0.76)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S2 (checkable).

As they design their digital product, complementors in platform ecosystems face a trade-off between value creation and value capture in their strategic decisions.

“As they design their digital product, complementors in platform ecosystems face a trade-off between value creation and value capture in their strategic decisions (Cennamo & Santaló, 2019;Jacobides et al., 2024).”

• cites Cennamo & Santaló, 2019 sourceNot checked — no full text of the source
• cites Jacobides et al., 2024 sourcePartly supported

  Section 3.3 (page not supplied) reports that videogame complementors underinvested in quality because they could not capture the value of their innovation. This connects product decisions, value creation, and value capture, but does not establish the claimed general trade-off faced by complementors when designing digital products. (LLM, conf=0.81)

  Experimental — accuracy not yet measured

Classified as probably not a substantial claim.

On the software level, complementors rely on platform-specific boundary resources, such as SDKs, that platforms provide to create ties to a particular ecosystem.

“On the software level, complementors rely on platform-specific boundary resources, such as SDKs, that platforms provide to create ties to a particular ecosystem (Engert et al., 2022;Zhao & Aaltonen, 2023).”

• cites Engert et al., 2022 sourceSupported
• cites Zhao & Aaltonen, 2023Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

Of all iOS gaming apps within the dataset, 54.79% were initially launched on the iOS platform.

“Of all iOS gaming apps within our dataset, 54.79% were initially launched on the iOS platform.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The bivariate correlations display a negative relationship between app Age and Multihoming (ß = -0.108; p < 0.001), indicating developers are more likely to choose multihoming with newer apps.

“The bivariate correlations display a negative relationship between our control variable app Age in months since initial release with Multihoming (ß = -0.108;p < 0.001), indicating that developers are more likely to choose multihoming with newer apps and vice versa.”

May combine two claims.

The relationship between Multihoming and Complement Quality is moderated in the absence of Direct Monetization (ß = 0.379; p < 0.001) and, less statistically significant, its presence (ß = 0.032; p < 0.050).

“The simple slope test indicates this relationship is moderated in case of the absence of Direct Monetization (ß = 0.379; p < 0.001) and, less statistically significant, its presence (ß = 0.032; p < 0.050).”

• AssumptionThe authors state this as an assumption: The dependent variable follows a gamma distribution with a logarithmic link function.

Probably not a substantial claim: may not pass S3 (load-bearing).

The goodness-of-fit of the GLM models, measured by deviance, continuously decreases, indicating an improving model fit of the hierarchical approach.

“To consider the goodness-of-fit of our GLM models, we rely on the deviance, which is continuously decreasing and thus indicates an improving model fit of our hierarchical approach.”

• AssumptionThe authors state this as an assumption: The dependent variable follows a gamma distribution with a logarithmic link function.

May combine two claims.

Propensity score matching estimates an average treatment effect on the treated apps (ß = 0.229; p < 0.001), suggesting that multihoming apps exhibit a higher average rating than singlehoming apps matched based on control variables.

“The estimated average treatment effect on the treated apps (ß = 0.229; p < 0.001) suggests that multihoming apps exhibit a higher average rating than singlehoming apps that we match based on the control variables.”

May combine two claims.

The interaction between Multihoming and Software Autonomy is significant only for higher levels of independence (ß = 0.178; p < 0.001) and not lower levels (ß = 0.031; p = 0.183).

“We estimate the marginal effect of the interaction with a simple slope test (Aiken et al., 1991) to confirm that this interaction term is significant only for higher levels of independence (ß = 0.178; p < 0.001) and not lower levels (ß = 0.031; p = 0.183)”

• cites Aiken et al., 1991Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• AssumptionThe authors state this as an assumption: The dependent variable follows a gamma distribution with a logarithmic link function.

The quote only partly states this claim.

In the GLM for iOS gaming apps, the interaction of multihoming and hardware autonomy has no significant association with complement quality (coefficient 0.007, n.s.).

“Multihoming × Hardware Autonomy | 0.007”

The quote only partly states this claim.

In the GLM for iOS gaming apps, software autonomy has no significant association with complement quality (coefficient 0, n.s.).

“Software Autonomy | 0”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

It remains understudied how established platform ecosystem dynamics such as multihoming affect value creation through the novel theoretical lens of data network effects.

“In particular, it remains understudied (1) how established platform ecosystem dynamics such as multihoming affect value creation through the novel theoretical lens of data network effects, (2) how data network effects extend to the complement-level through data-driven feedback loops, and (3) how complementors' decisions on value capture influence the process of value creation.”

Classified as probably not a substantial claim. May combine two claims.

Data network effects emerge from faster and more precise insights into consumer data by deploying AI capabilities in platform ecosystems, which may affect perceived consumer quality.

“Accordingly, data network effects emerge from faster and more precise insights into consumer data by deploying AI capabilities in platform ecosystems, which may affect perceived consumer quality (Clough & Wu, 2022;Gregory et al., 2021).”

• cites Clough & Wu, 2022 sourceNot checked — no full text of the source
• cites Gregory et al., 2021 sourceNot checked — no full text of the source

Classified as probably not a substantial claim.

Recent technological innovations, such as multiplatform SDKs or AI-based developer tools, have reduced the complexity of cross-platform application development.

“Recent technological innovations, such as multiplatform SDKs or AI-based developer tools, have reduced the complexity of cross-platform application development (Kang et al., 2019;Zhao & Aaltonen, 2023).”

• cites Kang et al., 2019Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• cites Zhao & Aaltonen, 2023Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The availability of complements significantly influences consumer value creation within the ecosystem, which is crucial for sustainable platform success.

“The availability of complements significantly influences consumer value creation within the ecosystem, which is crucial for sustainable platform success (Tian et al., 2022).”

• cites Tian et al., 2022 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

Multihoming complementors become part of a larger, more diverse group of peers whose activities might influence them.

“Third, multihoming complementors become part of a larger, more diverse group of peers whose activities might influence them (Cenamor, 2021;Srinivasan & Venkatraman, 2018).”

• cites Cenamor, 2021 sourceNot checked — no full text of the source
• cites Srinivasan & Venkatraman, 2018 sourceNot checked — no full text of the source

May combine two claims.

Multihoming grants complementors access to a wider user base and, accordingly, to significantly more and increasingly diverse user data feeding into AI models for data-driven learning.

“Multihoming grants complementors access to a wider user base (Tavalaei & Cennamo, 2021) and, accordingly, to significantly more and increasingly diverse user data feeding into AI models for data-driven learning.”

• Also stated in: § Conceptual model and hypotheses development · plain

  “By providing complements across multiple ecosystems, developers extend their potential market and gather data from a significantly larger and more diverse user base (Cennamo et al., 2018;Kang, 2022;Kang et al., 2019).”

• cites Tavalaei & Cennamo, 2021 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

It remains understudied how data network effects extend to the complement level through data-driven feedback loops.

“In particular, it remains understudied (1) how established platform ecosystem dynamics such as multihoming affect value creation through the novel theoretical lens of data network effects, (2) how data network effects extend to the complement-level through data-driven feedback loops, and (3) how complementors' decisions on value capture influence the process of value creation.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Complementors can incorporate AI functionalities, thereby facilitating data network effects of data-driven learning by choosing suitable SDKs.

“Complementors can also incorporate AI functionalities, thereby facilitating data network effects of data-driven learning by choosing suitable SDKs (Gregory et al., 2021), such as platform-agnostic tools to monitor user behavior (Unity Technologies, 2024).”

• cites Gregory et al., 2021 sourceNot checked — no full text of the source
• cites Unity Technologies, 2024Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

Regular quality updates positively affect complement performance.

“Regular quality updates positively affect complement performance (Lee & Raghu, 2014).”

• cites Lee & Raghu, 2014 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

Multihoming allows complementors to access a broader audience by operating on more than one platform.

“Second, multihoming allows complementors to access a broader audience by operating on more than one platform (Tavalaei & Cennamo, 2021).”

• cites Tavalaei & Cennamo, 2021 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

Digital platform ecosystems commonly draw on consumer ratings of product quality as a measure indicating the value the platform offering has for its users.

“Accordingly, digital platform ecosystems commonly draw on consumer ratings of product quality as a measure indicating the value the platform offering has for its users (Cennamo & Santaló, 2019;Kircher & Foerderer, 2024).”

• cites Cennamo & Santaló, 2019 sourceNot checked — no full text of the source
• cites Kircher & Foerderer, 2024 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

It remains understudied how complementors' decisions on value capture influence the process of value creation.

“In particular, it remains understudied (1) how established platform ecosystem dynamics such as multihoming affect value creation through the novel theoretical lens of data network effects, (2) how data network effects extend to the complement-level through data-driven feedback loops, and (3) how complementors' decisions on value capture influence the process of value creation.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Novel frameworks and toolkits for mobile app developers, such as Google's Flutter, allow the platform-agnostic development of complements based on a single codebase.

“Novel frameworks and toolkits for mobile app developers, for instance Google's Flutter, allow the platform-agnostic development of complements based on a single codebase (Google LLC, 2024).”

• cites Google LLC, 2024Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

Increased interaction on the same side of a network increases utility through economies of scale.

“Increased interaction on the same side of a network increases utility through economies of scale (Karhu et al., 2024;Katz & Shapiro, 1985)”

• cites Karhu et al., 2024 sourceNot checked — no full text of the source
• cites Katz & Shapiro, 1985Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

May combine two claims.

The model results with alternative operationalizations remain stable, inferring that findings are largely unaffected by measurement errors of key variables.

“As the model results with alternative operationalizations remain stable, we infer that our findings are largely unaffected by measurement errors of our key variables.”

• AssumptionThe authors state this as an assumption: The dependent variable follows a gamma distribution with a logarithmic link function.

Probably not a substantial claim: may not pass S3 (load-bearing).

Mobile device users tend to singlehome on one platform, meaning they use only one mobile device or one operating system.

“Mobile device users tend to singlehome on one platform, meaning they use only one mobile device or one operating system (Kang et al., 2019;Zhao & Aaltonen, 2023).”

• cites Kang et al., 2019Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• cites Zhao & Aaltonen, 2023Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

May combine two claims.

The potential to leverage user data affects complementors' strategic choices, including multihoming across different platforms, designing products with greater autonomy from the platform technology, or monetizing their product.

“The potential to leverage user data affects complementors' strategic choices, including multihoming across different platforms, designing products with greater autonomy from the platform technology, or monetizing their product.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Random sample checks on historic iOS app pages using the Internet Archive service were conducted to ensure data validity.

“Further, we conducted random sample checks on historic iOS app pages using the Internet Archive service to ensure data validity.”

Probably not a substantial claim: may not pass S3 (load-bearing).

To sustain performance by providing high consumer value, complementors depend on steady investments and financial resources for product innovation.

“Thus, to sustain performance by providing high consumer value, complementors depend on steady investments and financial resources for product innovation.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Platform researchers are interested in the impact of multihoming on quality.

“Thus, platform researchers are interested in the impact of multihoming on quality (Kang et al., 2019).”

• cites Kang et al., 2019Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

The study examines major endogeneity sources-omitted variables bias, sample selection bias, and potential measurement errors-to improve the robustness of model results.

“Additionally, we examine the major endogeneity sources-omitted variables bias, sample selection bias, and potential measurement errors-to improve the robustness of our model results.”

May combine two claims.

Data network effects are relevant not only for platform providers but also influence complementors' strategic considerations.

“Hence, data network effects are relevant not only for platform providers (Gregory et al., 2021) but also influence complementors' strategic considerations.”

• cites Gregory et al., 2021 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

The study fits the GLM in a hierarchical approach to ensure transparency on the contribution of explanatory variables to the variance of the dependent variable.

“To ensure transparency on the contribution of the explanatory variables to the variance of our dependent variable, we fit the GLM in a hierarchical approach.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The study addresses potential concerns regarding multicollinearity by following the three-step procedure by Kalnins (2018).

“By following the three-step procedure by Kalnins (2018), we address potential concerns regarding multicollinearity.”

• cites Kalnins (2018) sourceNot checked — no full text of the source

May combine two claims.

The inverse Mills ratio is statistically significant in the main model, indicating systematic differences between gaming and business apps that affect the model's results.

“The inverse Mills ratio is statistically significant in our main model, indicating systematic differences between gaming and business apps that affect our model's results.”

• AssumptionThe authors state this as an assumption: The dependent variable follows a gamma distribution with a logarithmic link function.

May combine two claims.

The instruments (Contains Violence and US-Operated) exhibit predictive validity for Multihoming (F = 18.49; p < 0.001) and are exogenous as both Sargan (p = 0.620) and Hausman (p = 0.193) tests are statistically insignificant.

“These instruments exhibit predictive validity for Multihoming (F = 18.49; p < 0.001) and are exogenous as both the Sargan (p = 0.620) and Hausman (p = 0.193) tests are statistically insignificant.”

• AssumptionThe authors state this as an assumption: The instruments (Contains Violence and US-Operated) affect the independent variable rather than the dependent variable.

Probably not a substantial claim: may not pass S3 (load-bearing).

Complementors aim to keep a certain independence from the dominant platform provider.

“Thus, complementors aim to keep a certain independence from the dominant platform provider (Cenamor, 2021;Srinivasan & Venkatraman, 2018).”

• cites Cenamor, 2021 sourceNot checked — no full text of the source
• cites Srinivasan & Venkatraman, 2018 sourceNot checked — no full text of the source

Classified as probably not a substantial claim.

Data completeness and consistency were ensured by focusing on mobile gaming apps ranked in the top 500 in the USA within the past 5 years.

“We ensured data completeness and consistency by focusing on mobile gaming apps ranked in the top 500 in the USA within the past 5 years.”

Classified as probably not a substantial claim.

Given the chosen context of the empirical study, the authors are limited to theorizing about the implications of data network effects on value creation reflected in increased complement quality.

“Second, given the chosen context of our empirical study, we are limited to theorizing about the implications of data network effects on value creation reflected in increased complement quality.”

• AssumptionRests on an assumption: Complement quality (average star rating of app) reflects consumer value creation.

Probably not a substantial claim: may not pass S3 (load-bearing).

A platform's AI capability leads to greater value as the more a platform learns from the data it collects on users, the more valuable it becomes to each user.

“A platform's AI capability leads to greater value as the more a platform learns from the data it collects on users, the more valuable it becomes to each user (Clough & Wu, 2022;Gregory et al., 2021)”

• cites Clough & Wu, 2022 sourceNot checked — no full text of the source
• cites Gregory et al., 2021 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

For multihoming complementors, accessing consumer data from various sources is particularly beneficial for targeted advertising.

“For multihoming complementors, accessing consumer data from various sources is particularly beneficial for targeted advertising (Ritala & Karhu, 2023).”

• cites Ritala & Karhu, 2023 sourceNot checked — no full text of the source

Classified as probably not a substantial claim.

Decoupling an app from the platform's hardware ensures platform changes do not directly affect the app.

“Decoupling an app from the platform's hardware ensures platform changes do not directly affect the app (Dellermann & Reck, 2017).”

• cites Dellermann & Reck, 2017 sourceNot checked — no full text of the source

May combine two claims.

While multihoming grants complementors access to more and increasingly diverse user data, greater software autonomy allows them to deploy SDKs to implement data network effects, independently of the platform provider's AI capabilities and interests.

“Thus, while multihoming grants complementors access to more and increasingly diverse user data, greater software autonomy allows them to deploy SDKs to implement data network effects (Gregory et al., 2021;TensorFlow, 2024), independently of the platform provider's AI capabilities and interests.”

• cites Gregory et al., 2021 sourceNot checked — no full text of the source
• cites TensorFlow, 2024Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

Increased variety of complements within a network increases utility through economies of scope.

“Increased variety of complements within a network increases utility through economies of scope (Karhu et al., 2024;Katz & Shapiro, 1985)”

• cites Karhu et al., 2024 sourceNot checked — no full text of the source
• cites Katz & Shapiro, 1985Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Classified as probably not a substantial claim.

Research indicates that overall consumer satisfaction decreases when gaming users need to purchase additional capabilities to increase their chances of winning, with the purchase request triggered by user-specific activities.

“For instance, research indicates that overall consumer satisfaction decreases when gaming users need to purchase additional capabilities to increase their chances of winning, with the purchase request triggered by user-specific activities (Clough & Wu, 2022).”

• cites Clough & Wu, 2022 sourceNot checked — no full text of the source

Classified as probably not a substantial claim.

Greater software independence from the focal platform helps developers realize the benefits of learning from others.

“Greater software independence from the focal platform helps developers realize the benefits of learning from others (Cenamor, 2021).”

• cites Cenamor, 2021 sourceNot checked — no full text of the source

Classified as probably not a substantial claim.

The multihoming complementor can translate faster and more precise predictions from user data into higher consumer value with data network effects without the risk of limited or already obsolete complement improvements.

“Hence, the multihoming complementor can translate faster and more precise predictions from user data into higher consumer value with data network effects (Gregory et al., 2021) without the risk of limited or already obsolete complement improvements.”

• cites Gregory et al., 2021 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

On a hardware level, complementors can leverage device-specific components supplied by the platform provider, such as the camera module or sensors of a mobile device.

“On a hardware level, complementors can leverage device-specific components supplied by the platform provider, such as the camera module or sensors of a mobile device (Yoo et al., 2010).”

• cites Yoo et al., 2010 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

A sufficiently large installed base of users affects products' utility for any user due to possible interaction.

“A sufficiently large installed base of users affects products' utility for any user due to possible interaction (Agarwal et al., 2023)”

• cites Agarwal et al., 2023Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S1 (asserted), S4 (based).

When interacting with other complementors (same-sided), complementors with lower dependence on a particular platform can leverage a greater variety of AI capabilities and deploy them to improve consumer value.

“When interacting with other complementors (same-sided), complementors with lower dependence on a particular platform can leverage a greater variety of AI capabilities and deploy them to improve consumer value.”

May combine two claims.

The paper's findings reveal a value-diminishing impact of both indirect and direct monetization measures in the context of multihoming.

“Our findings reveal a value-diminishing impact of both indirect and direct monetization measures in the context of multihoming.”

Probably not a substantial claim: may not pass S4 (based).

The positive cross-side data network effects from multihoming decrease if complementors leverage a broadened user base for data-driven learning and apply the insights to capture value.

“Hence, we argue that the positive cross-side data network effects from multihoming decrease if complementors leverage a broadened user base for data-driven learning and apply the insights to capture value.”

Classified as probably not a substantial claim.

For short, iterative innovation cycles, such as mobile apps' continuous version updates, a platform-agnostic complement design strengthens positive data network effects, enabling consumer value from multihoming.

“However, for short, iterative innovation cycles, such as mobile apps' continuous version updates, we argue that a platform-agnostic complement design strengthens positive data network effects, enabling consumer value from multihoming.”

Classified as probably not a substantial claim.

Multihoming effects on value creation depend on complement design and platform characteristics.

“Multihoming effects on value creation depend on complement design and platform characteristics (Tavalaei & Cennamo, 2021;Zhao & Aaltonen, 2023)”

• cites Tavalaei & Cennamo, 2021 sourceNot checked — no full text of the source
• cites Zhao & Aaltonen, 2023Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

The study investigates variance in complement quality using cross-sectional data, so it cannot capture data network effects that occur over time as feedback loops.

“First, while we investigate variance in complement quality using cross-sectional data, data network effects presumably occur over time as feedback loops (Clough & Wu, 2022;Karhu et al., 2024).”

• cites Clough & Wu, 2022 sourceNot checked — no full text of the source
• cites Karhu et al., 2024 sourceNot checked — no full text of the source

May combine two claims.

By decoupling apps from the underlying platform, developers are more independent of the platform provider and less affected by changes in the platform's architecture.

“By decoupling apps from the underlying platform, developers are more independent of the platform provider and less affected by changes in the platform's architecture (Dellermann & Reck, 2017).”

• cites Dellermann & Reck, 2017 sourceNot checked — no full text of the source